Welcome back to the Abstract! These are the studies this week that reached back through time, flooded the zone, counted the stars, scored science goals, and topped it all off with a ten-course meal.

First, scientists make a major breakthrough thanks to a very cute mammoth mummy. Then: the climate case for busy beavers; how to reconnect with 3,000 estranged siblings; this is your brain on football; and last, what Queen Elizabeth II had for lunch on February 20, 1957.

 As always, for more of my work, check out my book First Contact: The Story of Our Obsession with Aliens, or subscribe to my personal newsletter the BeX Files. 

The long afterlife of Yuka the mammoth

Mármol Sánchez, Emilio et al. “Ancient RNA expression profiles from the extinct woolly mammoth.” Cell.

Scientists have sequenced RNA—a key ingredient of life as we know it—from the remains of a mammoth that lived 39,000 years ago during the Pleistocene “Ice Age” period, making it by far the oldest RNA on record. 

The previous record holder for oldest RNA was sourced from a puppy that lived in Siberia 14,300 years ago. The new study has now pushed that timeline back by an extraordinary 25,000 years, opening a new window into ancient genetics and revealing a surprise about a famous mammoth mummy called Yuka. 

“Ancient DNA has revolutionized the study of extinct and extant organisms that lived up to 2 million years ago, enabling the reconstruction of genomes from multiple extinct species, as well as the ecosystems where they once thrived,” said researchers led by Emilio Mármol Sánchez of the Globe Institute in Copenhagen, who completed the study while at Stockholm University.

“However, current DNA sequencing techniques alone cannot directly provide insights into tissue identity, gene expression dynamics, or transcriptional regulation, as these are encoded in the RNA fraction.”

“Here, we report transcriptional profiles from 10 late Pleistocene woolly mammoths,” the team continued. “One of these, dated to be ∼39,000 years old, yielded sufficient detail to recover…the oldest ancient RNA sequences recorded to date.”

DNA, the double-stranded “blueprint” molecule that stores genetic information, is far sturdier than RNA, which is why it can be traced back for millions of years instead of thousands. Single-stranded RNA, a “messenger” molecule that carries out the orders of DNA, is more fragile and rare in the paleontological record.

In addition to proving that RNA can survive much longer than previously known, the team discovered that Yuka—the mammoth that died 39,000 years ago—has been misgendered for years (yes, I realize gender is a social construct that does not apply to extremely dead mammoths, but mis-sexed just doesn’t have the same ring). 

Yuka was originally deemed female according to a 2021 study that observed the “presence of skin folds in the genital area compatible with labia vulvae structures in modern elephants and the absence of male-specific muscle structures.” Mármol Sánchez and his colleagues have now overturned this anatomical judgement by probing the genetic remnants of Yuka’s Y chromosome.

In fact, as I write this on Thursday, November 13—a day before the embargo on this study lifts on Friday—Yuka is still listed as female on Wikipedia. 

Just a day until you can live your truth, buddy.

In other news…

Leave it to beavers 

Burgher, Jesse A. S. et al. “Beaver-related restoration and freshwater climate resilience across western North America.” Restoration Ecology.

Every era has a champion; in our warming world, eager beavers may rise to claim this lofty title. 

These enterprising rodents are textbook “ecosystem engineers” that reshape environments with sturdy dams that create biodiverse havens that are resistant to climate change. To better assess the role of beavers in the climate crisis, researchers reviewed the reported behavioral beaver-related restoration (BRR) projects across North America. 

“Climate change is projected to impact streamflow patterns in western North America, reducing aquatic habitat quantity and quality and harming native species, but BRR has the potential to ameliorate some of these impacts,” said researchers led by Jesse A. S. Burgher of Washington State University. 

The team reports “substantial evidence that BRR increases climate resiliency…by reducing summer water temperatures, increasing water storage, and enhancing floodplain connectivity” while also creating “fire-resistant habitat patches.” 

So go forth and get busy, beavers! May we survive this crisis in part through the skin of your teeth.

One big happy stellar family

Boyle, Andrew W. et al. “Lost Sisters Found: TESS and Gaia Reveal a Dissolving Pleiades Complex.” The Astrophysical Journal.

Visible from both the Northern and Southern Hemispheres, the Pleiades is the most widely recognized and culturally significant star cluster in the night sky. While this asterism is defined by a handful of especially radiant stars, known as the Seven Sisters, scientists have now tracked down thousands of other stellar siblings born from the same clutch scattered across some 2,000 light years.

“We find that the Pleiades constitutes the bound core of a much larger, coeval structure” and “we refer to this structure as the Greater Pleiades Complex,” said researchers led by Andrew W. Boyle of the University of North Carolina at Chapel Hill. “On the basis of uniform ages, coherent space velocities, detailed elemental abundances, and traceback histories, we conclude that most stars in this complex originated from the same giant molecular cloud.” 

The work “further cements the Pleiades as a cornerstone of stellar astrophysics” and adds new allure to a cluster that first exploded into the skies during the Cretaceous age. (For more on the Pleiades, check out this piece I wrote earlier this year about the deep roots of its lore).

Getting inside your head(er)

Zamorano, Francisco et al. “Brain Mechanisms across the Spectrum of Engagement in Football Fans: A Functional Neuroimaging Study.” Radiology.

Scientists have peered into a place I would never dare to visit—the minds of football fans during high-stakes plays. To tap into the neural side of fanaticism, researchers enlisted 60 healthy male fans from the ages of 20 to 45 to witness dozens of goal sequences from matches involving their favorite teams, rival teams, and “neutral” teams while their brains were scanned by an fMRI machine. 

The participants were rated according to a “Football Supporters Fanaticism Scale (FSFS)” with criteria like “violent thought and/or action tendencies” and “institutional belonging and/or identification.” The scale divided the group up into 38 casual spectators, 19 committed fans, and four deranged fanatics (adjectives are mine for flourish).

“Our key findings revealed that scoring against rivals activated the reward system…while conceding to rivals triggered the mentalization network and inhibited the dorsal anterior cingulate cortex (dACC)”—a region responsible for cognitive control and decision-making—said researchers led by Francisco Zamorano of the Universidad San Sebastián in Chile. “Higher Football Supporters Fanaticism Scale scores correlated with reduced dACC activation during defeats, suggesting impaired emotional regulation in highly engaged fans.”

In other words, it is now scientifically confirmed that football fanatics are Messi bitches who love drama. 

Diplomacy served up fresh

Cabral, Óscar et al “Power for dinner. Culinary diplomacy and geopolitical aspects in Portuguese diplomatic tables (1910-2023).”

We’ll close, as all things should, with a century of fine Portuguese dining. In yet another edition of “yes, this can be a job,” researchers collected 457 menus served at various diplomatic meals in Portugal from 1910 to 2023 to probe “how Portuguese gastronomic culture has been leveraged as a culinary diplomacy and geopolitical rapprochement strategy.” 

As a lover of both food and geopolitical bureaucracy, this study really hit the spot. Highlights include a 1957 “regional lunch” for Queen Elizabeth II that aimed to channel “Portugality” through dishes like lobster and fruit tarts from the cities of Peniche and Alcobaça. The study is also filled with amazing asides like “the inclusion of imperial ice cream in the European Free Trade Association official luncheon (ID45, 1960) seems to transmit a sense of geopolitical greatness and vast governing capacity.” Ice cream just tastes so much better when it’s a symbol of international power. 

The team also unearthed a possible faux pas: Indian president Ramaswamy Venkataraman, a vegetarian who was raised Hindu, was served roast beef in 1990. In a footnote, Cabral and his colleagues concluded that “further investigation is deemed necessary to understand the context of ‘roast beef’ service to the Indian President in 1990.” Talk about juicy gossip!

 Thanks for reading! See you next week.

Every month or so, Anthropic puts out a press release about how we should all be very frightened!! of AI. This month’s scare story is a Chinese hacker scare! [Anthropic; report, PDF]

Anthropic says that someone who “we assess with high confidence was a Chinese state-sponsored group” has been using the agentic abilities of chatbots and automating their hacking runs. And they used Anthropic’s Claude Code to do it!

We believe this is the first documented case of a large-scale cyberattack executed without substantial human intervention.

“Without substantial human intervention” there means, by Anthropic’s estimates, “10 to 20 percent of total effort.”

Can you guess what Anthropic’s advice is? I bet you can!

We advise security teams to experiment with applying AI for defense in areas like Security Operations Center automation, threat detection, vulnerability assessment, and incident response.

As if security teams don’t already use all the machine learning AI they can.

There’s a few things in this report you might question straight away.

If it’s a Chinese state actor … why are they using Claude Code? Why not Chinese chatbots like DeepSeek or Qwen? Those chatbots code just about as well as Claude. Anthropic do not address this really obvious question.

Attackers automate their attacks already. They don’t type every command in live. That’s where we get the phrase “script kiddie” from — someone running someone else’s hacking script.

Chatbots are already used in attacks. The Nx hack was a pretty clever hack that used a chatbot — where the attacker got the victim to run their code, and it called any local coding bots on the victim’s PC to steal the victim’s cryptocurrency. The main security revelation from Nx was “wow, vibe coders really are idiots.”

Anthropic’s most implausible claim is that an AI agent did it. Agents just do not work reliably. You are not going to get a chatbot to reliably automate a long attack chain.

Dan Tentler of Phobos tells Ars Technica: [Ars Technica]

I continue to refuse to believe that attackers are somehow able to get these models to jump through hoops that nobody else can. Why do the models give these attackers what they want 90% of the time but the rest of us have to deal with ass-kissing, stonewalling, and acid trips?

Kevin Beaumont notes: [Mastodon; Mastodon]

There’s no IoCs [indicators of compromise] at all in the report, or the usual threat indicators. The threats it describes are all already widely detected.

They’re showing vibe usage of open source attack frameworks — which should be fine for existing detections and controls. The threat actors aren’t inventing something new here.

Whatever Anthropic detected, they spent two months writing up this report — and didn’t ask actual security guys at any point.

Plus, China! Gotta get those defense dollars!

This report is marketing for Anthropic and Claude Code, and you shouldn’t take it seriously.

• Video — Podcast

Click here to go see the bonus panel!

Hovertext:
Anyone claiming this comic has any actual perspective will be accused of being in league with JSquatch.

Pre-orders for my new book Sawyer Lee and the Quest to Just Stay Home have begun!

Sawyer Lee is an illustrated middle grade novel starring an unadventurous kid who'd rather dig a deep dent in the couch than make a mark on the world, as many in his illustrious family of astronauts, scientists, spies, champion athletes... blah blah blah... have. He has decided that after generations of effort, it’s time to spend one lifetime relaxing. 

The problem is that Sawyer keeps getting caught up in the exhausting expectations of his wicked aunt Celia, his complex relationship with his ambitious other friend, Angela, and the shenanigans of every else in town hoping to win the yearly Gourd Thump festival celebrating nature’s dullest vegetable.

In this tale of mystery, treachery, conspiracy, plant husbandry, and an imaginary love triangle, Sawyer knows it will take a regrettable amount of energy to escape these entanglements and find a way back to his happy place on Gary’s couch, with a cozy throw blanket, a steaming mug of chamomile tea, and an empty schedule.

You can check out the first chapter here along with pre-order links!

Microsoft and nuclear power company Westinghouse Nuclear want to use AI to speed up the construction of new nuclear power plants in the United States. According to a report from think tank AI Now, this push could lead to disaster. 

“If these initiatives continue to be pursued, their lack of safety may lead not only to catastrophic nuclear consequences, but also to an irreversible distrust within public perception of nuclear technologies that may inhibit the support of the nuclear sector as part of our global decarbonization efforts in the future,” the report said.

The construction of a nuclear plant involves a long legal and regulatory process called licensing that’s aimed at minimizing the risks of irradiating the public. Licensing is complicated and expensive but it’s also largely worked and nuclear accidents in the US are uncommon. But AI is driving a demand for energy and new players, mostly tech companies like Microsoft, are entering the nuclear field. 

“Licensing is the single biggest bottleneck for getting new projects online,” a slide from a Microsoft presentation about using generative AI to fast track nuclear construction said. “10 years and $100 [million.]”

The presentation, which is archived on the website for the US Nuclear Regulatory Commission (the independent government agency that’s charged with setting standards for reactors and keeping the public safe), detailed how the company would use AI to speed up licensing. In the company’s conception, existing nuclear licensing documents and data about nuclear sites data would be used to train an LLM that’s then used to generate documents to speed up the process.

 But the authors of the report from AI Now told 404 Media that they have major concerns about trusting nuclear safety to an LLM. “Nuclear licensing is a process, it’s not a set of documents,”  Heidy Khlaaf, the head AI scientist at the AI Now Institute and a co-author of the report, told 404 Media. “Which I think is the first flag in seeing proposals by Microsoft. They don’t understand what it means to have nuclear licensing.”

“Please draft a full Environmental Review for new project with these details,” Microsoft’s presentation imagines as a possible prompt for an AI licensing program. The AI would then send the completed draft to a human for review, who would use Copilot in a Word doc for “review and refinement.” At the end of Microsoft’s imagined process, it would have “Licensing documents created with reduced cost and time.”

The Idaho National Laboratory, a Department of Energy run nuclear lab, is already using Microsoft’s AI to “streamline” nuclear licensing. “INL will generate the engineering and safety analysis reports that are required to be submitted for construction permits and operating licenses for nuclear power plants,” INL said in a press release. Lloyd's Register, a UK-based maritime organization, is doing the same. American power company Westinghouse is marketing its own AI, called bertha, that promises to make the licensing process go from "months to minutes.”

The authors of the AI Now report worry that using AI to speed up the licensing process will bypass safety checks and lead to disaster. “Producing these highly structured licensing documents is not this box taking exercise as implied by these generative AI proposals that we're seeing,” Khlaaf told 404 Media. “The whole point of the lesson in process is to reason and understand the safety of the plant and to also use that process to explore the trade offs between the different approaches, the architectures, the safety designs, and to communicate to a regulator why that plant is safe. So when you use AI, it's not going to support these objectives, because it is not a set of documents or agreements, which I think you know, is kind of the myth that is now being put forward by these proposals.”

Sofia Guerra, Khlaaf’s co-author, agreed. Guerra is a career nuclear safety expert who has advised the U.S. Nuclear Regulatory Commission (NRC) and works with the International Atomic Energy Agency (IAEA) on the safe deployment of AI in nuclear applications. “This is really missing the point of licensing,” Guerra said of the push to use AI. “The licensing process is not perfect. It takes a long time and there’s a lot of iterations. Not everything is perfectly useful and targeted …but I think the process of doing that, in a way, is really the objective.”

Both Guerra and Khlaaf are proponents of nuclear energy, but worry that the proliferation of LLMs, the fast tracking of nuclear licenses, and the AI-driven push to build more plants is dangerous. “Nuclear energy is safe. It is safe, as we use it. But it’s safe because we make it safe and it’s safe because we spend a lot of time doing the licensing and we spend a lot of time learning from the things that go wrong and understanding where it went wrong and we try to address it next time,” Guerra said.

Law is another profession where people have attempted to use AI to streamline the process of writing complicated and involved technical documents. It hasn’t gone well. Lawyers who’ve attempted to write legal briefs have been caught, over and over again, in court. AI-constructed legal arguments cite precedents that do not exist, hallucinate cases, and generally foul up legal proceedings.

Might something similar happen if AI was used in nuclear licensing? “It could be something as simple as software and hardware version control,” Khlaaf said. “Typically in nuclear equipment, the supply chain is incredibly rigorous. Every component, every part, even when it was manufactured is accounted for. Large language models make these really minute mistakes that are hard to track. If you are off in the software version by a letter or a number, that can lead to a misunderstanding of which software version you have, what it entails, the expectation of the behavior of both the software and the hardware and from there, it can cascade into a much larger accident.”

Khlaaf pointed to Three Mile Island as an example of an entirely human-made accident that AI may replicate. The accident was a partial nuclear meltdown of a Pennsylvania reactor in 1979. “What happened is that you had some equipment failure and design flaws, and the operators misunderstood what those were due to a combination of a lack of training…that they did not have the correct indicators in their operating room,” Khlaaf said. “So it was an accident that was caused by a number of relatively minor equipment failures that cascaded. So you can imagine, if something this minor cascades quite easily, and you use a large language model and have a very small mistake in your design.”

In addition to the safety concerns, Khlaaf and Guerra told 404 Media that using sensitive nuclear data to train AI models increases the risk of nuclear proliferation. They pointed out that Microsoft is asking not only for historical NRC data but for real-time and project specific data. “This is a signal that AI providers are asking for nuclear secrets,” Khlaaf said. “To build a nuclear plant there is actually a lot of know-how that is not public knowledge…what’s available publicly versus what’s required to build a plant requires a lot of nuclear secrets that are not in the public domain.”

“This is a signal that AI providers are asking for nuclear secrets. To build a nuclear plant there is actually a lot of know-how that is not public knowledge…what’s available publicly versus what’s required to build a plant requires a lot of nuclear secrets that are not in the public domain.”

Tech companies maintain cloud servers that comply with federal regulations around secrecy and are sold to the US government. Anthropic and the National Nuclear Security Administration traded information across an Amazon Top Secret cloud server during a recent collaboration, and it’s likely that Microsoft and others would do something similar. Microsoft’s presentation on nuclear licensing references its own Azure Government cloud servers and notes that it’s compliant with Department of Energy regulations. 404 Media reached out to both Westinghouse Nuclear and Microsoft for this story. Microsoft declined to comment and Westinghouse did not respond.

“Where is this data going to end up and who is going to have the knowledge?” Guerra told 404 Media.

Nuclear is a dual use technology. You can use the knowledge of nuclear reactors to build a power plant or you can use it to build a nuclear weapon. The line between nukes for peace and nukes for war is porous. “The knowledge is analogous," Khlaaf said. “This is why we have very strict export controls, not just for the transfer of nuclear material but nuclear data.”

Proliferation concerns around nuclear energy are real. Fear that a nuclear energy program would become a nuclear weapons program was the justification the Trump administration used to bomb Iran earlier this year. And as part of the rush to produce more nuclear reactors and create infrastructure for AI, the White House has said it will begin selling old weapon-grade plutonium to the private sector for use in nuclear reactors.

Trump’s done a lot to make it easier for companies to build new nuclear reactors and use AI for licensing. The AI Now report pointed to a May 23, 2025 executive order that seeks to overhaul the NRC. The EO called for the NRC to reform its culture, reform its structure, and consult with the Pentagon and the Department of Energy as it navigated changing standards. The goal of the EO is to speed up the construction of reactors and get through the licensing process faster.

A different May 23 executive order made it clear why the White House wants to overhaul the NRC. “Advanced computing infrastructure for artificial intelligence (AI) capabilities and other mission capability resources at military and national security installations and national laboratories demands reliable, high-density power sources that cannot be disrupted by external threats or grid failures,” it said.

At the same time, the Department of Government Efficiency (DOGE) has gutted the NRC.  In September, members of the NRC told Congress they were worried they’d be fired if they didn’t approve nuclear reactor designs favored by the administration. “I think on any given day, I could be fired by the administration for reasons unknown,” Bradley Crowell, a commissioner at the NRC said in Congressional testimony. He also warned that DOGE driven staffing cuts would make it impossible to increase the construction of nuclear reactors while maintaining safety standards. 

“The executive orders push the AI message. We’re not just seeing this idea of the rollback of nuclear regulation because we’re suddenly very excited about nuclear energy. We’re seeing it being done in service of AI,” Khlaaf said. “When you're looking at this rolling back of Nuclear Regulation and also this monopolization of nuclear energy to explicitly power AI, this raises a lot of serious concerns about whether the risk associated with nuclear facilities, in combination with the sort of these initiatives can be justified if they're not to the benefit of civil energy consumption.”

Matthew Wald, an independent nuclear energy analyst and former New York Times science journalist is more bullish on the use of AI in the nuclear energy field. Like Khlaaf, he also referenced the accident at Three Mile Island. “The tragedy of Three Mile Island was there was a badly designed control room, badly trained operators, and there was a control room indication that was very easy to misunderstand, and they misunderstood it, and it turned out that the same event had begun at another reactor. It was almost identical in Ohio, but that information was never shared, and the guys in Pennsylvania didn't know about it, so they wrecked a reactor,” Wald told 404 Media.

"AI is helpful, but let’s not get messianic about it.”

According to Wald, using AI to consolidate government databases full of nuclear regulatory information could have prevented that. “If you've got AI that can take data from one plant or from a set of plants, and it can arrange and organize that data in a way that's helpful to other plants, that's good news,” he said. “It could be good for safety. It could also just be good for efficiency. And certainly in licensing, it would be more efficient for both the licensee and the regulator if they had a clearer idea of precedent, of relevant other data.”

He also said that the nuclear industry is full of safety-minded engineers who triple check everything. “One of the virtues of people in this business is they are challenging and inquisitive and they want to check things. Whether or not they use computers as a tool, they’re still challenging and inquisitive and want to check things,” he said. “And I think anybody who uses AI unquestionably is asking for trouble, and I think the industry knows that…AI is helpful, but let’s not get messianic about it.”

But Khlaaf and Guerra are worried that the framing of nuclear power as a national security concern and the embrace of AI to speed up construction will setback the embrace of nuclear power. If nuclear isn’t safe, it’s not worth doing. “People seem to have lost sight of why nuclear regulation and safety thresholds exist to begin with. And the reason why nuclear risks, or civilian nuclear risk, were ever justified, was due to the capacity for nuclear power. To provide flexible civilian energy demands at low cost emissions in line with climate targets,” Khlaaf said.

“So when you move away from that…and you pull in the AI arms race into this cost benefit justification for risk proportionality, it leads government to sort of over index on these unproven benefits of AI as a reason to have nuclear risk, which ultimately undermines the risks of ionizing radiation to the general population, and also the increased risk of nuclear proliferation, which happens if you were to use AI like large language models in the licensing process.”

Sometimes a song just comes out of nowhere and does great! But also, there’s well worn paths to buying your way to a small amount of public attention. And these ways have been common as long as there’s been mass media with music in. Payola is the least of it.

Spotify plays and followers are a commodity. You can just buy them. We wrote up last year how a guy bought himself a ton of Spotify streams of AI tracks. Instagram followers are a commodity you can just buy.

And if someone cracks out Suno and generates yet another song going “the lights are low but the beat is high,” they’re not going to get organic attention.

But if they can get gullible idiots in the music press to report on them, they’re much happier. And there are journalists who will write up anything that says a robot might replace humans.

We’ve seen the fake fashion version of this a couple of times, where a brand promises things that are actually impossible, or some chancers say they’re using AI because they’re trying to sell you a PDF on how to make money like them. These would not have worked without foolish journalists who fail to ask the most basic and obvious questions.

I am blaming the journalists who write up these fakes as if they are not the fakes they really obviously are. And you should blame the journalists too.

It’s the same with fake bands. I am reluctant to even mention this variety of scoundrel, because even Pivot to AI levels of attention would just feed their hype machine. The bozo behind the Velvet Sundown project even sent me a press release about his latest AI fake band project, saying I should expose it. Then he signed his name.

But the fake bands do suck in some gullible press fools who write like marketing never happened. I’m looking at you, Brian Hiatt from Rolling Stone, making Velvet Sundown look serious to the rest of the press. You have no excuse not to have known better. [Rolling Stone]

The key tell is that all the numbers are ones that are really easy to fake. You can buy followers and clicks. You can do fake deals where you claim a huge headline number but no actual money moves. You can’t buy organic interest.

Nobody talks about the bands except the obviously AI-generated comments. Nobody is interested, except they might click on it once because there was a headline in an easily-fooled news outlet. There is no evidence anyone actually likes AI music slop.

Today we have a “band” called Breaking Rust. It topped the Billboard country charts! Wow! Well, it topped the Billboard country digital song sales chart. That’s one of the very tiny also-ran charts.

Breaking Rust’s total sales to top the country downloads chart was … 3,000 downloads. They bought their way onto the chart for $3,000. [Independent]

That was enough for gullible media, including a gullible writer at Billboard itself — Xander Zellner is our lucky loser today, as the guy who had no excuse not to know better: [Billboard]

at least one AI artist has debuted in each of the past six chart weeks, a streak suggesting this trend is quickly accelerating.

Maybe the “trend” of buying yourself coverage. I know music journalism is a super tough game right now. But that was just a very dumb thing to write.

When you see one of these AI bands or AI singers or AI actors, check for anything verifiable and real. Check for numbers that can’t be faked. You won’t find them. Because nobody wants AI slop artists.

You will find easily fooled writers who really need to step away from the keyboard and think for a bit. It’s not that hard! C’mon guys, you can do it! Give it a go!

• Video — Podcast

Click here to go see the bonus panel!

Hovertext:
Are 23 eggs not an oeuf for you? (Pun brought to you by Patreon comments)